INDEX

Explanations

pointed shape, sharp edges, tapered, angular

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

å·¦æīĭ

-0.22

é¦Ļåĳ³

-0.21

æ³īæ°´

-0.20

æ»ĭ

-0.20

æĭ¿äºĨ

-0.20

åĳ»

-0.19

 trem

-0.19

è¹Ĵ

-0.19

æ¼«æŃ¥

-0.18

å¹¸ç¦ıæĦŁ

-0.18

POSITIVE LOGITS

å°ĸ

0.29

 sharper

0.24

ness

0.22

éĶĲ

0.21

 sharp

0.20

å½¢

0.19

åŀĭ

0.19

 trunc

0.19

corner

0.18

éĻ·

0.18

Activations Density 0.023%